AI 中文总结
本研究开发嵌入亚纳米PtNPs的有机晶体管，利用类库 block 输运实现大存储窗口，兼具长短时程可塑性，为神经形态计算中STP的实现提供新路径。
AI 中文摘要
有机晶体管在神经形态应用中发挥着日益重要的作用。然而，仅依赖铁电开关或界面陷阱实现多电导态的器件，其存储窗口有限。本研究在薄膜晶体管结构中，于聚合物半导体与铁电介质的界面处引入含均匀分布亚纳米铂纳米颗粒（PtNPs）的超薄氧化层；该界面堆叠结构Al₂O₃/PtNP/Al₂O₃为局域电荷的捕获与去捕获提供了可行途径，且该区域可最有效地影响沟道电导。所制备的有机晶体管在电流-电压特性中展现出大于20 V的大存储窗口，亚纳米PtNPs产生的特性与室温类库 block 输运一致，支持存储窗口内离散且界限分明的能级。该器件可通过电与光刺激实现多模态编程，兼具长时程可塑性与增强型短时程可塑性（STP）现象，为神经形态计算发展中STP的实现开辟了新方向。
英文摘要
Organic transistors are playing an increasingly important role for neuromorphic applications. However, devices that rely solely on ferroelectric switching or on interface traps for their multi-conductance states exhibit limited memory windows. Here, we introduce an ultrathin oxide layer with a uniform distribution of sub-nm platinum nanoparticles (PtNPs) at the interface of a polymer semiconducting and a ferroelectric dielectric in a thin film transistor architecture. The interfacial stack, Al$_2$O$_3$/PtNP/Al$_2$O$_3$, provides a viable route for localized charge trapping and de-trapping in a region where it can most effectively influence the channel conductance. The organic transistors display a large memory window (> 20 V) in their current-voltage characteristics. The sub-nm PtNPs give rise to features that are consistent with room temperature Coulomb blockade-like transport, supporting discrete and well-separated levels within the memory window. The devices support multimodal programming using electrical and optical stimuli with both long-term plasticity and enhanced short-term plasticity (STP) phenomena. These results open new directions for implementing STP in the development of neuromorphic computing.
Comments19 pages, 4 figures plus supplemental information (7 pages, 8 figures)
